Terrestrial Carbon Cycle - Fingerprints of Changes in Response to Large Ocean Circulation Reorganisation

Australia: The Land Where Time Began

In this paper Bozbiyik et al. present the results of their investigation into changes in the land, ocean and atmosphere CO₂and carbon by use of the comprehensive carbon cycle-climate model NCAR CSM1.4-carbon. Freshwater perturbations are applied to deep water formation sites in the North Atlantic and Southern Ocean under pre-industrial climate conditions forcing ensemble simulations. This results in the Atlantic Meridional Overturning Circulation (AMOC) being reduced by varying amounts in each experiment. There is a clear distinction between the perturbations in the Northern and Southern Hemispheres, the physical climate fields showing changes that are qualitatively in agreement with results that have been documented in the literature. The biogeochemical cycles, both terrestrial and oceanic, are in turn affected by the changes in the physical variables, and cause either an increase or decrease in the atmospheric concentration of CO₂ of up to 20 ppmv, depending on the location of the perturbation. When the perturbation is in the North Atlantic the land biosphere reacts in some tropical locations and at high northern latitudes with a strong reduction of carbon stocks, though these stocks tend to increase in response to a perturbation in the Southern Hemisphere. The ocean is usually a carbon sink, though throughout various basins large reorganisations occur. The land biosphere responds most strongly in tropical regions, as a result of the Intertropical Convergence Zone being shifted. In South America the carbon fingerprint of this shift can be seen most clearly, whether the shift is to the north or the south, depending on the location where the freshwater is applied. As a result of this a compilation of various proxy records of precipitation changes in the Younger Dryas are compared with the results from the Bozbiyik et al. model. The response to a freshwater perturbation in the North Atlantic shows that the proxy records are generally in good agreement with the response of the model.

It is suggested by records from various climate proxies, especially those in Greenland ice cores, as well as sediments from the North Atlantic, that large, abrupt climatic changes have occurred during the last glacial period (Stocker, 2000; Rahmstorf, 2002; Clement & Peterson, 2008). Those transitions occurred on time scales of as little as a few years to a few decades. Ice cores from North Grip, Greenland, contain evidence of 26 such abrupt warming events that have local amplitudes of up to 16^oC (NorthGRIP Members, 2004: Huber et al., 2006; Steffensen et al., 2008). These are known as Dansgaard-Oeschger (D-O) events (Dansgaard et al., 1984; Oeschger et al., 1984). There are also periods of intense cold that accompany warm events, preceding these events for which corresponding signals have been found in the sediment cores from the northern Atlantic, which are marked by distinct layers of ice rafted debris (Heinrich, 1988; Bond et al., 1993; Hemming, 2004). Heinrich events are surges of large amounts of ice into the sea, which are associated with some of the coldest temperatures in Greenland. Also, it is indicated by high-resolution measurements on Antarctic ice cores that each of the Dansgaard-Oeschger (D-O) events has an abrupt counterpart in the Southern Hemisphere (EPICA Community Members, 2006), that is likely to due to the bipolar seesaw (Stocker & Johnsen, 2003). The Younger Dryas cold event is the last event in this sequence, and it terminated with an abrupt warming (D-O 0). There are also signatures of those climatic events that have been found in other climate archives, such as the isotopic and pollen records in lake and marine sediments (Eicher et al., 198; Ruddiman & McIntyre, 1981; Bond et al., 1983; Yu & Eicher, 1998; Ammann et al., 2000; Baker et al., 2001; Prokopenko et al., 2001; Voelker, 2002; Hughen et al., 2004; Barker et al., 2009) and European and Asian loess records (Ding et al., 1999; Porter et al., 2001; Rousseau et al., 2002) with a geographic distribution from Europe to Asia and beyond.

It has been found that many of the characteristics of an abrupt climate change event can be simulated by climate models that are coupled, that are perturbed by anomalous fluxes of freshwater applied in the North Atlantic (Bryan et al., 1986; Mikolajewicz, 1996; Schiller et al., 1997; Manabe & Stouffer, 1999; Marchal et al., 1998; Timmermann et al., 2003; Knutti et al., 2004; Zhang & Delworth, 2005; Stocker & Marchal, 2000; Stouffer et al., 2006), which is capable of causing the Atlantic Meridional Overturning Circulation (AMOC) to collapse.

The aim of this paper was to determine the dependence of the climate and carbon cycle to the freshwater perturbations of different origins and to identify the fingerprints of these responses by the use of a comprehensive atmosphere-ocean circulation model coupled with a land surface model. Also, the changes in in the carbon cycle during a collapse of the AMOC under pre-industrial conditions, and considering northern as well southern origins. In this respect, the ocean response and the terrestrial biosphere response have been investigated focusing especially on a region of South America where the model simulates responses that are particularly strong. The reasons for these abrupt changes that are observed in the palaeoclimate records have been surmised previously in 1984, as a result of the nonlinear nature of the ocean-climate system (Oeschger et al., 1984). The existence of different modes of thermos-haline circulation is 1 such nonlinearity in the system. At least some of the abrupt climate changes that occurred in the past, which include the Younger Dryas event which resulted in intense cold around the northern Atlantic, as well as having climate impacts around the globe, are considered to result from such rapid reorganisations (Boyle & Keigwin, 1987; Duplessy et al., 1988; Broecker, 1997; Clark et al., 2002). The slowing of the ventilation of the ocean in the North Atlantic during these cold periods is documented in the records of ¹⁴C and ¹⁰Be (Hughen et al., 2000; Muscheler et al., 2000).

According to Bozbiyik et al. any interruption in the ocean-wide circulation would result in result in climatic effects, on global scales as well as local scales. Included among those effects are intense cooling of the Northern Hemisphere, which is centred around Northern Europe and Greenland which then spreads to the northern Pacific (Okumura et al., 2009); changes in marine ecosystems in the Atlantic (Schmittner, 2005) and seas level in the North Atlantic (Levermann et al., 2005); precipitation pattern changes over the tropics due to the Inter-Tropical Convergence Zone (ITCZ) (Vellinga & Wood, 2002; Dahl et al., 2005) and changes in the El Niño-Southern Oscillation phenomenon (Timmermann et al., 2005, 2007). The tropics play an important role in the abrupt climate change events, by globalising the shutdown of the AMOC, which is a Northern Hemisphere phenomenon, by reorganisation of the ocean and the atmosphere (Chiang, 2009).

In order to explain the changes in concentration of atmospheric CO₂ during those abrupt events it is necessary to understand the response of the global carbon cycle to a large input of freshwater into the ocean. At times of intense cooling events in Greenland, and the more gradual warm events in Antarctica, records of atmospheric CO₂ show small though significant variations. Atmospheric concentrations of CO₂ rose by 20 ppmv during Antarctic warm events A1 to A4 (Stauffer et al., 1998; Indermühle et al., 2000), and during the much shorter Younger Dryas cold event, by about the same amount (Monnin et al., 2001).

In regard to the source of this increase there are 2 different ideas, either the ocean or the changes in vegetation on land. Some have suggested that the increase in atmospheric CO₂ was the result of the release of carbon on the land (Kӧhler et al., 2005; Obata, 2007; Menviel et al., 2008), though it has been suggested by some modelling experiments that the increase in atmospheric CO₂ resulted from the release of carbon on land (Marchal et al,. 1999; Schmittner & Galbraith, 2008). An explanation for increasing CO₂ concentrations can be ocean outgassing if the cooling of the surface of the ocean is constrained to the northern latitudes, the warming of the Southern Ocean is more pronounced and the land contribution is not taken into account (Marchal et al., 1999). The ocean was also identified as a source for increase of atmospheric CO₂ during abrupt climate changes events (Schmittner & Galbraith, 2008). Their model is probably limited in representing changes in tropical precipitation that have impacts that are potentially large on the biosphere of the land, as a result of the absence of a complex atmospheric component.

The land based biosphere is the other possible contributor to changes in the carbon cycle that occurred during abrupt climate change events. It has been found (Kӧhler et al., 2005) that under pre-industrial conditions and pre-Younger Dryas conditions concentrations of atmospheric CO₂ rises as a result of the release of carbon from the land, with the latter being slightly less pronounced.

More recently a general circulation model coupled with a simple land surface model was employed (Obata, 2007) to simulate a shutdown of the AMOC which caused carbon to be released from the land with the result that there was an increase in atmospheric CO₂. It was suggested by another study using an Earth system model of intermediate complexity (LOVECLIM) (Menviel et al., 2008) that when the AMOC was shut down the ocean acted as a carbon sink and the land as a carbon source under both pre-industrial and Last Glacial Maximum (LGM) conditions. The results of these model studies are supported in many ways by the results of this study by Bozbiyik et al. whose study extended it further by providing a clearer picture of the biosphere on land. A feature of the study by Bozbiyik et al. which gives insight as to which hemisphere might have triggered such events in the past is an alternative location for the application of freshwater perturbations. The fingerprint of each trigger, whether Northern Hemisphere of Southern Hemisphere, are evident in the South American continent, which is the centre of action for terrestrial carbon cycle changes. A comparison of the results of the model by Bozbiyik et al. with palaeorecords in the region is also provided.

Discussion and conclusions

The response of the climate and the collapse of the AMOC, as indicated by the model of Bozbiyik et al., can be divided into 2 categories, depending on where the freshwater perturbation is applied. All the perturbations from the region of formation of North Atlantic Deep Water cause similar responses, which are opposite to those caused by perturbations being applied from the Weddell Sea and the Ross Sea. The responses of the Weddell Sea and the Ross Sea also differ from each other, with with the Ross Sea causing a much more widespread cooling and a precipitation signal that is clearer. Bozbiyik et al. suggest that seems to be because the Ross Sea appears to a more important player in the formation of AABW in the model of Bozbiyik et al. and, therefore, the perturbations in the Ross Sea has a stronger effect on the global climate. It is noted by Bozbiyik et al., however, that the specific location of the formation of deep water are model specific.

Around the tropics near the ITCZ is the location of the most significant precipitation changes occurring and its position is sensitive to sea surface temperature (SST) shifts. This, in turn, leads to large changes in the carbon stocks in these locations. Also, these precipitation anomalies are caused to amplify changes in carbon stocks by the fact that large amounts of carbon stocks are stored in low latitudes. The Lund-Potsdam-Jena (LPJ) model was forced with an output from freshwater experiments with the ECBILT-CLIO model Kӧhler et al., 2005). Large carbon stock changes in the tropics in the boreal zone were found and in the tropics, relatively small carbon stock changes, in contrast to the results obtained by Bozbiyik et al. According to Bozbiyik et al. the 2 studies are different in a number of ways. There are a variety of differences between the 2 studies. In the LPJ vegetation dynamics is explicitly simulated, whereas the distribution of vegetation is prescribed in the NCAR C SM1.4 carbon model.

Interactions and feedbacks between vegetation and climate, such as those that are related to albedo and the water cycle, on the other hand, are represented in the coupled NCAR model, though not in the forced runs with LPJ. The ability to simulate the dynamics in the tropics is limited by the atmospheric dynamics being represented in a simplified manner in the cost-efficient ECBILT-CLIO.

Following the northern perturbations the southwards shift of the ITCZ is well studied by the use climate models and is supported by palaeoclimatic records (Leduc et al., 2009), it is shown by Bozbiyik et al. that the opposite is also true, i.e., the shift to the north of the ITCZ in response to a southern perturbation. The direction of the change in concentration of atmospheric CO₂ through its effects on land carbon pool in the low latitudes is determined by the direction of this shift.

During the Heinrich events, as well as during the Younger Dryas event, the magnitude of the increase in CO₂ was around 20 ppmv, as has been observed in ice cores (Indermühle et al., 2000; Monnin et al., 2001). According to the results obtained by Bozbiyik et al. the carbon release from the land biosphere can explain all of this amplitude while the oceans act as a carbon sink. An origin on land for the increase in CO₂ during the Younger Dryas is indicated by the isotopic signature of the CO₂ from the ice cores (Smith et al., 1999). The major features of the Heinrich Event 1 (H1) were produced successfully by a recent comprehensive simulation of the last deglaciation, covering the H1, with a coupled atmosphere-ocean general circulation model (Liu et al., 2009). They study by Liu et al. found strong cooling in the Northern Hemisphere, and a milder warming in the south, and reduced precipitation in the Cariaco Basin in northern South America. It is indicated by the good agreement of the general patterns of change that a comparison of a palaeoclimate event under glacial conditions (H1) and the simulation of Liu et al. under pre-industrial conditions are reasonable.

In the experiments by Liu et al. the net atmospheric CO₂ increase is comparable to, though not more than, that obtained with similar initial conditions (Obata, 2007). Though the changes in climatic variables and the NPP show strong resemblances their magnitudes differ in some areas. Included among these is a cooling that was more widespread in the Northern Hemisphere in the experiments carried out by Bozbiyik et al., and a smaller response by land biosphere to the precipitation anomalies in eastern Asia. There is also a larger negative anomaly recorded in the northern part of South America in the study of Bozbiyik et al. Yet, the general effect of the shift of the ITCZ is robust in both studies. Bozbiyik et al. suggest the differences may be attributed to the land biosphere representation in the model of Obata et al. (2007) being more limited.

The exact amount of the contribution of CO₂ increase made by the terrestrial biosphere to the atmosphere should, however, be taken with caution as there was a different vegetation cover on the land at times of glaciation was different than the one that was implemented here. The experiments that were carried out by Bozbiyik et al. were under pre-industrial conditions with a larger vegetation carbon pool than would have been the case during glacial times. In Experiments have previously been carried out (Menviel et al., 2008), under both pre-industrial and glacial boundary conditions, it was shown that the differences in the amplitude of individual contributions the carbon pools of the land and the ocean may lead to an opposite net effect on the levels of CO₂ in the atmosphere, in spite of the nature of each contribution is qualitatively the same. Irrespective of the initial state, in their study the roles of the ocean as a carbon sink and of the land as a carbon source remain unchanged. Also, there are very similar changes on land in both cases, i.e., reduced carbon stocks in the high and middle latitudes of the Northern Hemisphere and in the tropics to the north of the equator and an increase in the south. Yet, under glacial conditions the emissions from the land are weaker than under the pre-industrial conditions. That probably results from the lower moisture content of the glacial atmosphere, which leads to the dampening effects of the shift of the ITCZ (Menviel et al.,. 2008), and gains in primary production that are relatively large in some regions such as eastern Asia and southern North America. Compared with the experiments of Bozbiyik et al., are the larger increases in carbon stocks in the Southern Hemisphere, as well as the above mentioned regions of the north, which Bozbiyik et al. suggest might be the result of some model specific differences as well as the initial conditions.

According to Bozbiyik et al. it is safe to assume that their results are relevant for palaeoreconstructions as a possible indirect method to distinguish between the sources of the discharge of freshwater in abrupt cooling events as the patterns of anomalies are very similar in both studies, because the North Atlantic as well as the Antarctic perturbations have distinct implications for the biosphere on land. On a global scale, the atmospheric CO₂ signal is different; in the case of the North Atlantic an increase, and in the Antarctic case a decrease. Regionally, the South American continent has proven to be in a particularly suitable position to record such events in the past, as movement of the ITCZ – to either the north or the south – would form distinguishable and, at some places, opposite responses.

As well as what has been presented in the results section of this paper, there are also some other notable features that are observed in response to freshwater perturbations, such as the formation of the North Pacific Deep Water, or the strengthening of the Southern Hemisphere westerlies. A more detailed investigation of these responses may be undertaken in a future study.

Also, the results of this study have implications for anthropogenic climate change in the future, as has been shown by many modelling studies, is to cause a reduction in the AMOC (Meehlet et al., 2007). There can be substantial effects for climate and for low latitude ecosystems including, but not limited to, rainforests, of such a reduction. It is also important that the carbon cycle changes during such an event would possibly contribute to the increase in the atmospheric carbon and, therefore, operate as a weak positive feedback to the global warming, in addition to that associated with outgassing from a warmer ocean (Joos et al., 1999).

See Source 1 for more detailed information

Sources & Further reading